# Fileset

[Nagata_paper_2025.pdf](https://mdr.nims.go.jp/filesets/541a0fb4-8b0b-4fd8-8238-7e9ed88c4af6/download)

## Creator

Kanaka Iwasaki, Yuka Takagi, Hyunwook Nam, Hajime Nagata, [Isao Sakaguchi](https://orcid.org/0000-0003-4382-2509)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Electrical properties and <sup>18</sup>O tracer diffusion in (Bi<sub>0.5</sub>Na<sub>0.5</sub>)TiO<sub>3</sub> ceramics doped with CuO and Nb<sub>2</sub>O<sub>5</sub>](https://mdr.nims.go.jp/datasets/04c8eebc-d85c-47d4-8337-a710af763623)

## Fulltext

Electrical properties and 18O tracer diffusion in (Bi0.5Na0.5)TiO3 ceramics doped with CuO and Nb2O5FULL PAPERElectrical properties and 18O tracer diffusion in (Bi0.5Na0.5)TiO3ceramics doped with CuO and Nb2O5Kanaka Iwasaki1, Yuka Takagi1,³, Hyunwook Nam1, Hajime Nagata1 and Isao Sakaguchi21Tokyo University of Science, 2641 Yamazaki, Noda, Chiba 278–8510, Japan2National Institute for Materials Science, 1–1 Namiki, Tsukuba, Ibaraki 305–0044, Japan(Bi0.5Na0.5)TiO3 [BNT] ceramics can be sintered at low temperatures of 940 °C with CuO addition. This isbelieved to reduce Bi3+ volatilization and improve the quality of BNT-based ceramics by lowering the number ofoxygen vacancies. However, there is no direct evidence that low-temperature sintering prevents the formation ofoxygen vacancies. In this study, the formation of oxygen vacancies by 18O tracer diffusion and the additive effectof CuO on this formation in BNT ceramics were examined using secondary ion mass spectrometry (SIMS). Inaddition, Nb2O5, which acts as a donor, was added to the CuO-doped BNT ceramics. The formation of oxygenvacancies was discussed, and the electrical properties were clarified. As a result, the volume diffusion coefficientsD of the BNT ceramics with 0.5wt% CuO (Cu0.5) were 2.0 © 10¹11 cm2/s. This value is equivalent to pure BNT,which means that for BNT ceramics, CuO acts as an acceptor, suggesting that BNT ceramics still contain manyoxygen vacancies. On the other hand, the D of 18O in Cu0.5 added 0.4wt% Nb2O5 was 3.5 © 10¹15 cm2/s,indicating that the formation of oxygen vacancies is suppressed. Moreover, Nb2O5 enhanced the poling treatmentand the coercive field Ec decreased, indicating a softening trend.Key-words : Lead-free piezoelectric ceramics, Bi0.5Na0.5TiO3, Nb2O5 donor additive, 18O tracer diffusion[Received January 7, 2025; Accepted April 18, 2025; Published online May 20, 2025]1. IntroductionIn recent years, lead-free piezoelectric materials havegarnered considerable attention as environmentally friend-ly substitutes for lead-containing materials. Perovskite-structured lead-free piezoelectric materials, in particular,are well-suited for use in actuators and high-power de-vices.1,2) (Bi0.5Na0.5)TiO3 (BNT) ceramics are consideredpromising lead-free piezoelectric materials due to theirrelatively low production costs and excellent piezoelectriccharacteristics.3–16) Generally, BNT ceramics are known tohave the highest relative density when sintered at 1140 °C.However, secondary ion mass spectrometry (SIMS) analy-sis has revealed significant Bi3+ ion volatilization near1100 °C, raising concerns about potential compositionalinstability.17) Therefore, the BNT ceramics must be sin-tered at low temperatures. However, their d33 values arerelatively smaller compared to those of PZT-based ceram-ics. To address this limitation, incorporating multilayerstructures has proven highly effective for actuators andhigh-power devices, as the overall displacement scaleswith the number of layers. In the case of (K,Na)NbO3-based ceramics, the multilayer ceramic actuators (MLCAs)utilizing inner Ni electrodes have demonstrated outstand-ing piezoelectric performance.18,19) However, Bi-basedceramic MLCAs cannot be manufactured using base metalelectrodes like Ni. This is because the partial pressure ofoxygen (pO2) levels for the Bi/Bi2O3 system fall withinthe range of 10¹10 to 10¹12 atm, which are consistentlyhigher than those of the Ni/NiO system across all temper-atures, as indicated by the Ellingham diagram.20) In otherwords, Bi-based oxide ceramics are more easily reducedthan Ni electrodes. Therefore, the ceramics cannot be sin-tered in a reducing atmosphere without the Ni oxidation.Consequently, the fabrication of Bi-based MLCAs requiresusing precious metals such as Ag, Pt, and Ag–Pd.Typically, Ag–Pd electrodes are commonly utilized inPZT-based MLCAs due to their cost-effectiveness. Themelting point of Ag–Pd alloys decreases as the proportionof costly Pd is reduced.21) Recently, a 95:5 Ag–Pd ratio hasbeen employed in the production of PZT-based MLCAs,necessitating a sintering temperature below 990 °C. Sim-ilarly, Ag–Pd electrodes have also been used as internalelectrodes in fabricating MLCAs from BNT-based ce-ramics.22) At co-sintering temperatures of around 1100–1130 °C, the Ag–Pd electrodes were not fully activated,leading to a reduction in the electrical performance of theMLCA compared to bulk ceramics. Furthermore, the diffu-sion behavior of Ag used as an electrode material wasanalyzed through SIMS.23) The diffusion of Ag at the grainboundaries significantly affected the chemical and physical³ Corresponding author: Y. Takagi; E-mail: y-takagi@rs.tus.ac.jp‡ Preface for this article: DOI https://doi.org/10.2109/jcersj2.133.P7-1Journal of the Ceramic Society of Japan 133 [7] 321-328 2025DOI https://doi.org/10.2109/jcersj2.25012 JCS-Japan©2025 The Ceramic Society of Japan 321This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/),which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.https://doi.org/10.2109/jcersj2.133.P7-1https://doi.org/10.2109/jcersj2.133.P7-1https://doi.org/10.2109/jcersj2.25012https://creativecommons.org/licenses/by/4.0/strengths of the grain boundaries of the BNT ceramics. Thediffusion coefficient at the grain boundaries in BNT ceram-ics is comparable to that of Pb-based ceramics.24,25) Thesereports indicate that low-temperature sintering of BNTbased ceramics is essential for the co-sintering of internalelectrodes and BNT-based ceramics.CuO is commonly employed as a sintering aid owing toits low melting point and ability to form a liquid–phase inboth BNT-based ceramics26–40) and (K,Na)NbO3 (KNN)-based ceramics.41–47) The presence of the liquid–phasegreatly improves mass transport along grain boundaries,leading to increased final densities. In particular, sinter-ing aids can diffuse into ceramics and negatively affectedtheir properties.48) Moreover, an excessive amount of CuOmay lead to abnormal grain growth driven by the liquid–phase,49) which can further deteriorate piezoelectric perfor-mance,50,51) similar to what is observed in KNN. Therefore,it is essential to minimize the concentration of sinteringaids. In addition, effective densification at low temperaturesrequires strong adhesion between the liquid–phase andceramic particles. A high surface tension is required be-tween the liquid–phase and the particles, which is reflectedby a low contact angle.52) However, determining the con-tact angle of the liquid–phase is challenging, as it must bemeasured experimentally at the sintering temperature.53) Inour previous study, we prepared BNT ceramics doped with1.0wt% CuO (BNT + CuO), and clarified their electricalproperties and low-temperature sintering mechanisms. Therole of CuO were clarified as causing the formation of aliquid-phase of compounds of Cu, Na, Ti, and O ions(NCT). The sintering of BNT ceramics was achieved at940 °C, which is 200 °C lower than 1140 °C.54) Low tem-perature sintering at 940 °C by CuO doping can suppressthe Bi vaporization during the sintering, leading to thesuppression of formation of oxygen vacancies. However,no direct results have been reported showing that low-temperature sintering suppresses the formation of oxygenvacancies. Moreover, the same average grain size of theBNT + CuO as that of pure BNT was realized at a sinter-ing temperature 200 °C lower than that of the pure BNTceramics. Thus, it is assumed that the addition of CuOpromoted grain growth. This grain growth is attributed tothe enhanced diffusion of oxygen vacancies. When theTi4+-site of BNT perovskite structure is replaced by Cu2+ions, they have the potential to act as acceptors. However,the role of CuO as a dopant and the number of oxygenvacancies in BNT + CuO have not been clarified. Further-more, the concentration of oxygen vacancies, as measuredby 18O tracers using SIMS, decreases in BNTceramics withthe addition of 0.4wt% Nb2O5.14) This reduction is attrib-uted to the substitution of Ti4+ sites by Nb5+ in the BNTperovskite structure. Furthermore, the addition of Nb2O5increases the resistivity of BNT ceramics, while at the sametime reducing the coercive force (Ec), showing a softeningtendency. In other words, Nb2O5 effectively suppresses theformation of oxygen vacancies and acts as a donor in BNTceramics.14) Therefore, if Cu2+ ions are acting as acceptorsin BNT ceramics, it is expected that co-doping with Nb5+ions will be effective in suppressing the formation ofoxygen vacancies. In this study, we examined the formationof oxygen vacancies via 18O tracer diffusion in CuO-dopedBNT ceramics using secondary ion mass spectrometry(SIMS). In addition, the number of oxygen vacancies inBNT ceramics with co-doped, CuO and Nb2O5, was dis-cussed using 18O tracer diffusion, and the electrical prop-erties were evaluated.2. Experimental(Bi0.5Na0.5)TiO3 + xCuO (x = 0, 0.5, 1.0wt%, Cux)and BNT + Cux + yNb2O5 (y = 0, 0.2, 0.4, 0.8, 1.0wt%,Cux + Nby) ceramics were prepared using conventionalceramic fabrication techniques. The starting materials weremixed in ethanol via ball milling with zirconia balls for20 h. After drying, the mixed powder was calcined in analumina crucible at 850 °C for 2 h. Nb2O5 was added to thecalcined powder of Cux. The calcined powder was re-ground via ball milling for 20 h, then pressed into pelletswith a diameter of 20mm and a thickness of 10mm usinga uniaxial press. The pellets were subsequently treatedwith cold isostatic pressing at 150MPa. Each sample wassintered at 940 °C for 2 h. The ceramic densities weredetermined using Archimedes method. The crystal struc-ture and lattice parameters of the bulk samples wereanalyzed using X-ray diffraction (XRD, Rigaku RINT-2000, 40 kV, 40mA) with Cu-K¡ radiation. By using theresults of XRD, the lattice distortion 90-¡ was obtained.Here, the alpha (¡) value was determined by numericallysolving the following equation.1d2¼ðh2 þ k2 þ l2Þ sin2 ¡þ 2ðhk þ klþ hlÞðcos2 ¡� cos¡Þa2ð1� 3 cos2 ¡þ 2 cos3 ¡Þð1ÞIn this equation, d is the interplanar spacing measured withXRD, and (h, k, l) are Miller indices.An Ag electrode was applied to the ceramic surface formeasuring its piezoelectric properties as well as the resis-tivity. The resistivity was determined by RA/t, where Rwas the resistance measured with High-Resistance Meter(YHP 4339B) made by Keysight Technologies, Inc., and Aand t are the area and thickness of the electrode that has thediameter of 10mm and the thickness of 0.5mm. Longi-tudinal vibration in the (33) mode was evaluated usingrectangular specimens with dimensions 2mm © 2mm ©5mm. The samples were poled in a silicone oil bath undera DC electric field of 5 kV/mm for 5min at room tem-perature, then assessed using the resonance-antiresonancemethod as per IEEE standards with an impedance analyzer(HP4294A). The P-E hysteresis loops were measuredusing a virtual ground system (Toyo Corporation, Model6252 Rev. C) at room temperature and 10Hz. The 18Oconcentration profile was analyzed using a SIMS system(Cameca IMS-4F) at the National Institute for MaterialsScience, Japan. A 133Cs+ beam accelerated at 10 kV wasused as the primary ion to irradiate the sample surface. TheIwasaki et al.: Electrical properties and 18O tracer diffusion in (Bi0.5Na0.5)TiO3 ceramics doped with CuO and Nb2O5JCS-Japan322primary beam was scanned in a raster pattern over a 100 ©100¯m2 area. A Cameca normal-induced electron gun wasused to stabilize the surface potential during 133Cs+ irra-diation. A dynamic transfer system defined the measure-ment area of secondary ions and the raster window rang-ing from 40 to 50%. The secondary ions, 16O¹ (referencemass), and 18O¹, from this area were detected using anelectron multiplier. The 18O tracers were diffused byannealing the samples at 450 °C for 30min in an 18Oatmosphere. The concentration profile was calculated fromthe oxygen intensities as follows:C ¼ Ið18OÞIð18OÞ þ Ið16OÞ ð2Þwhere I(16O) and I(18O) are the intensities of 16O and 18Opeaks, respectively. The volume diffusion contribution inthe profile was fitted to the diffusion equation for a con-stant concentration at the surface as follows:Cx � CbgCs � Cbg¼ erfcx2ffiffiffiffiffiffiDtp� �ð3Þwhere Cx is the 18O concentration at depth x, Cs is thesurface concentration of 18O, and Cbg is the natural back-ground abundance of 18O. Equation (2) assumes an equi-librium between the crystal surface and gas phase and thusa constant Cs value. D and t are the volume diffusioncoefficient and annealing time, respectively.3. Results and discussion3.1 Effect of CuO on oxygen vacanciesformationFigure 1 shows the 18O tracer diffusion profiles ofCu0.5, which were compared with the profiles of BNTceramics with no additives (pure BNT), MgO 0.4wt%(Mg0.4), and Nb2O5 0.4wt% (Nb0.4). The volume diffu-sion coefficients D of Cu0.5 were 2.0 © 10¹11 cm2/s. Thisvalue is equivalent to that of pure BNT, 2.5 © 10¹11 cm2/s.BNT ceramics were originally p-type and have recentlybecome known as ionic conductors,55) which means thatmany oxygen vacancies are present in BNT ceramics.Thus, Cu0.5 was assumed to contain many oxygen vacan-cies. Moreover, in our previous study, the D of Mg0.4, andNb0.4 exhibits 9.2 © 10¹11 and 1.1 © 10¹13 cm2/s, respec-tively.14) It is suggested that this large difference in D val-ues is due to substituting of Ti4+ sites by Mg2+ and Nb5+ions, i.e., the difference in the number of oxygen vacanciesas 18O tracer diffusion. The additives MgO and Nb2O5 actas an acceptor and donor, for BNT ceramics.17) This meansthat the value of D can indicate of the amount of oxygenvacancies. The D value of the BNT ceramics with Cu0.5was similar to that of the BNT ceramics with Mg0.4. Thus,CuO behaved as an acceptor for BNT ceramics becauseCu2+ ions replaced the Ti4+ sites. As shown in our previ-ous study, Cu0.5 can be sintered at a low temperature of940 °C.16) The low-temperature sintered CuO-doped BNTceramics are believed to suppress the volatilization of Bi3+ions from around 1100 °C, thus avoiding the formation ofoxygen vacancies. However, CuO-doped BNT ceramicscontain large numbers of oxygen vacancies even when sin-tered at low temperatures. Based on these results, Nb2O5,is added to the CuO-doped BNT ceramics to suppress theformation of oxygen vacancies. In the next section, wediscuss the oxygen diffusion and electrical properties ofthe CuO-doped BNT ceramics with added Nb2O5.3.2 CuO-doped BNT ceramics with addedNb2O5Figure 2 shows the XRD patterns of Pure BNT, Cu0.5,Cu1.0, Nb0.4, Cu0.5 + Nb0.4, and Cu1.0 + Nb0.4. TheXRD pattern shows a rhombohedral perovskite structure,with a detailed observation showing a peak at approx-imately 34°. This peak was attributed to the liquid-phaseNCT formed by adding of CuO. The relative density ofCux + Nby (x = 0.5, 1.0wt%, y = 0, 0.2, 0.4, 0.8, 1.0wt%) averaged 98% as shown in Fig. 3(a). The latticeconstants for these samples averaged 3.89¡, and thesevalues did not change significantly for any CuO and Nb2O5content as shown in Fig. 3(b). Furthermore, Fig. 3(c)shows the lattice distortion 90-¡ as a function of theamount of Nb2O5 in y phase. The 90-¡ of the samples didnot change significantly until the addition of 0.2wt%Nb2O5. On the other hand, the 90-¡ value of the samplesFig. 1. 18O tracer diffusion profiles of pure BNT, Cu0.5,Mg0.4, and Nb0.4 after annealing at 450 °C for 30min.Fig. 2. XRD patterns of Pure BNT, Cu0.5, Cu1.0, Nb0.4,Cu0.5 + Nb0.4, and Cu1.0 + Nb0.4. (The peak at approximately34° was attributed to the liquid-phase NCT formed by adding ofCuO).Journal of the Ceramic Society of Japan 133 [7] 321-328 2025 JCS-Japan323with 0.4% or more of Nb2O5 added decreased slightly.However, these changes are not sufficiently pronouncedto cause a change in the lattice constant, as shown inFig. 3(b).Figure 4(a) shows the SEM image of the grains. Themapping of each element, such as Na, Bi, Ti, Cu, and Nbions, is shown in Figs. 4(b)–4(f ) and Fig. 4(g) is an over-lay of the elemental maps. A Cu-based grain was iden-tified at the triple points of the grains. While Na+, Ti4+,and Nb5+ ions were found within the Cu-based grains, ashighlighted by white dashed circles, Bi ions were absent.Furthermore, in Cu-based grains, grains with and withoutTi ions were observed as shown in Fig. 4(g). Hence, it isproposed that the Cu-based grains consist of compoundscontaining Cu, Na, Ti, Nb, and O ions instead of only Cu.A closer inspection of the XRD pattern in Fig. 2 revealsa peak near 34°. This peak closely resembles that ofNaCu2.5Ti4.5O12 (PDF No. 01-078-5412, referred to asNCT).54) It has been documented that compounds contain-ing Cu and Ti,56) as well as those with Cu and Na57), canlower the melting point. The melting point of Cu-basedgrain was therefore reduced when combined with Na andTi. This indicates that NCT forms a liquid–phase duringsintering due to its lowered melting point. Furthermore, theaddition of Nb suggests that Nb is also incorporated intothe liquid-phase NCT. Therefore, it is suggested that thereare both Nb5+ ions contained in the liquid–phase and thosethat replace the B-site of BNT.Figure 5 shows the 18O tracer diffusion profile ofCu0.5 + Nb0.4, which was compared to those of pureBNT and Cu0.5. The D of 18O in Cu0.5 + Nb0.4 was3.5 © 10¹15 cm2/s and this value is slower than that ofCu0.5. This indicates that the addition of the donor Nb5+ions suppressed the formation of oxygen vacancies inCu0.5. Low-temperature sintering and reduced oxygenvacancies were achieved for Cu0.5 + Nb0.4. In addition,the SEM images of the grains of Cu0.5 and Cu0.5 +Nb0.4 are shown in Figs. 6(a)–6(b) and 6(c)–6(d), respec-tively. The average grain sizes of these were 5.53 and 2.74¯m, respectively. As shown in previous reports, the grainFig. 3. (a) Relative densities, (b) Lattice constants a, and (c) lattice distortion 90-¡ for pure BNT, Cux + Nby(x = 0.5, 1.0wt%, y = 0, 0.2, 0.4, 0.8, 1.0wt%).Fig. 4. (a) SEM image of the grains, (b)–(f ) mapping of each element, such as Na, Bi, Ti, Cu, and Nb ions, and(g) overlay of the elemental maps.Iwasaki et al.: Electrical properties and 18O tracer diffusion in (Bi0.5Na0.5)TiO3 ceramics doped with CuO and Nb2O5JCS-Japan324size of Cu0.5 was larger than that of pure BNT ceramics(4.18¯m).54) This is because the grain grew due to thediffusion of oxygen vacancies, as shown in Fig. 5. In con-trast, for Cu0.5 + Nb0.4, the formation of oxygen vacan-cies is suppressed, and as a result, grain growth is inhib-ited. The average grain sizes of Cu0.5 and Cu1.0 withrespect to the amount of Nb2O5 added is summarized inFig. 6(c). The average grain size decreased as the amountof Nb2O5 added increased. Therefore, the grain growth ob-servation also showed that adding of Nb suppressed theformation of oxygen vacancies.To determine the optimum amount of Nb2O5, theamount of Nb2O5 added was controlled for the Cu0.5 andCu1.0. Figure 7 shows resistivity µ as a function of dopedthe Nb2O5 amount of y. The µ showed a maximum valueof 1013³ cm at y = 0.4. This value is comparable for bothCu0.5 and Cu1.0. Since BNT is originally p-type, theaddition of the donor Nb2O5 caused the charge to approachneutrality, thus increasing µ. Previously, a trend of in-creasing resistivity was observed for Nb0.4.14) For valuesof x greater than 0.4, the µ decreased. This suggested thatthe number of Nb5+ ions was too high to act effectively asdonors. Thus, approximately 0.4wt% Nb2O5 was assumedto be optimal in this system.The P-E hysteresis loops of pure BNT, Cu1.0 andCu1.0 + Nb0.4 at room temperature and 10Hz wereshown in Fig. 8(a). The loops were fully saturated andexhibited good rectangularities. The coercive electric fieldEc values of pure BNT, Cu1.0, and Cu1.0 + Nb0.4 were74.8, 75.7, and 69.7 kV/cm, respectively. The Ec of Cu1.0was not significantly different from that of pure BNT,while the Ec of Cu1.0 + Nb0.4 was lower than those ofCu1.0 and pure BNT. Figure 8(b) shows the Ec of Cu1.0with respect to the amount of Nb2O5 added. Ec decreasedas the amount of Nb2O5 added increased. In other words, asoftening trend is observed with the addition of Nb2O5,indicating that the formation of oxygen vacancies is sup-pressed. This is also consistent with the fact that CuOaddition alone does not suppress the formation of oxygenvacancies, as shown in Fig. 1. The softening trend shownin Fig. 8 is also consistent with the relationship betweenthe softening (or hardening) trend and the decrease (orincrease) in the oxygen vacancies shown in Ref. 14. In thisreference,14) the softening trend induced with the additionof only Nb2O5 without CuO addition. In the same ref-erence,14) the hardening trend with the addition of onlyMgO is shown to be induced by the increase in oxygenvacancies.Fig. 5. 18O tracer diffusion profile of pure BNT, Cu0.5, andCu0.5 + Nb0.4 after annealing at 450 °C for 30min.Fig. 6. SEM images of grain structures and average grain sizes for (a), (b) Cu0.5 and (c), (d) Cu0.5 + Nb0.4.(e) Average grain sizes of Cux + Nby (x = 0.5, 1.0wt%) as a function of the amount of Nb2O5 y.Fig. 7. Resistivity µ of Cu0.5 and Cu1.0 as a function of theamount of Nb2O5 y.Journal of the Ceramic Society of Japan 133 [7] 321-328 2025 JCS-Japan325Figure 9 shows the piezoelectric properties (a) the elec-tromechanical coupling factor k33, (b) the piezoelectricconstant d33 (c) the relative free permittivity ¾33T/¾0, (d)the elastic compliance s33E, and (e) the mechanical qualityfactor Qm as functions of the amount of Nb2O5 y. k33 showsa maximum value of 0.43 at y = 0.4 for both Cu0.5 andCu1.0, which is almost the same as the value for pureBNT. Also, d33 shows a maximum value of 83.6 pC/N aty = 0.4 for Cu0.5. d33 increased at y = 0.4 compared toy = 0 for Cu0.5 and Cu1.0. This is because, the additionof Nb2O5 reduces oxygen vacancies as shown in Fig. 5;therefore, the domain pinning effect is smaller, and polar-ization inversion is easier. ¾33T/¾0 are maintained atapproximately 400, independent of the amount of Nb2O5added. For Cu1.0 + Nb0.8, the values of ¾33T/¾0 showsapproximately 600 which means that ªmax was 76°, whichindicating that polarization inversion did not proceed.Moreover, s33E increased and Qm decreased with the addi-tion of Nb2O5 addition as shown in Figs. 9(d) and 9(e),respectively. This indicates that the samples are softenedFig. 8. (a) P-E hysteresis loops were compared among pure BNT, Cu1.0 and Cu1.0 + Nb0.4 at roomtemperature and 10Hz. (b) Relationship between Ec and the amount of Nb2O5 y.Fig. 9. Piezoelectric properties (a) k33, (b) d33, (c) ¾33T/¾0, (d) s33E, and (e) Qm of Cu0.5 and Cu1.0 as functionsof the amount of Nb2O5 y.Iwasaki et al.: Electrical properties and 18O tracer diffusion in (Bi0.5Na0.5)TiO3 ceramics doped with CuO and Nb2O5JCS-Japan326both materially and electrically, and this is due to the sup-pression of the formation of oxygen vacancies. In otherwords, the domain pinning effect of the oxygen vacancieswas small.Figure 10 shows the depolarization temperature Td ofCux + Nby (x = 0.5, 1.0wt%, y = 0, 0.2, 0.4, 0.8, 1.0wt%) as a function of the amount of Nb2O5. In the case ofy = 0, the addition of Cu causes the Td to decrease byabout 10 °C compared to pure BNT which is consistentwith the previous report.54) Furthermore, the addition ofNb tended to reduced Td further. The tendency of Td todecrease with the addition of Nb2O5 is consistent with aprevious report.14) However, the Td of Cu0.5 sintered at940 °C is around 170 °C, while the Td increases to around190 °C due to the quenching effect.8–16) Thus, applying thequenching effect makes it possible to increase Td of Cux +Nby.4. ConclusionWhile the addition of only Nb2O5 showed no effect ofsuppressing Bi3+ volatilization with only the effect of sup-pressing oxygen vacancies in the previous studies, theaddition of only CuO showed little effect of suppressingoxygen vacancies with only the effect of suppressing Bi3+volatilization due to the lowered sintering temperature. Inthis study, the number of oxygen vacancies for 18O tracerdiffusion was evaluated by SIMS for CuO-doped BNTceramics with Nb2O5 as a donor additive; subsequentlytheir electrical properties were examined. Consequently, inCuO-doped BNT ceramics with Nb2O5, oxygen vacanciesand Bi3+ volatilization were simultaneously suppressed.The addition of Nb2O5 resulted in softening tendency andimproved the piezoelectric properties. These approachescan guide fabricating high-quality BNT ceramics with sup-pressed oxygen vacancies.References1) T. Takenaka, H. Nagata, Y. Hiruma and K. Matsumoto,J. Electroceram. 19, 259 (2007).2) S. E. Park and T. R. Shrout, IEEE T. Ultrason. Ferr. 44,1141 (1997).3) T. Takenaka and H. Nagata, J. Eur. Ceram. Soc. 12,2693 (2005).4) H. Nagata, N. Koizumi, N. Kuroda, I. Igarashi and T.Takenaka, Ferroelectrics 229, 273 (1999).5) L. Liu and H. Fan, J. Electroceram. 16, 293 (2006).6) H. Nagata, J. Ceram. Soc. Jpn. 116, 271 (2008).7) T. Hoshina, Y. Kigoshi, T. Furuta, H. Takeda and T.Tsurumi, Jpn. J. Appl. Phys. 50, 09NC07 (2011).8) H. Muramatsu, H. Nagata and T. Takenaka, Jpn. J. Appl.Phys. 55, 10TB07 (2016).9) T. Miura, H. Nagata and T. Takenaka, Jpn. J. Appl.Phys. 56, 10PD05 (2017).10) H. Nagata, Y. Takagi, Y. Yoneda and T. Takenaka, Appl.Phys. Express 13, 061002 (2020).11) Y. Takagi, T. Miura, H. Nagata and T. Takenaka, Jpn. J.Appl. Phys. 58, SLLD02 (2019).12) Y. Takagi, H. Nagata and T. Takenaka, J. Asian Ceram.Soc. 8, 277 (2020).13) K. Eguchi, Y. Takagi, H. Nagata and T. Takenaka, Jpn.J. Appl. Phys. 59, SPPD03 (2020).14) Y. Takagi, K. Eguchi, H. Nagata, I. Sakaguchi and T.Takenaka, J. Ceram. Soc. Jpn. 129, 383 (2021).15) Y. Takagi, Y. Ochiai, M. Ito, T. Kawagoe, H. Nagata andI. Sakaguchi, Jpn. J. Appl. Phys. 61, SN1034 (2022).16) Y. Takagi, Y. Ochiai and H. Nagata, Jpn. J. Appl. Phys.60, SFFD02 (2021).17) H. Nagata, T. Watanabe, Y. Hiruma and T. Takenaka,Ferroelectrics 404, 2732 (2010).18) S. Kawada, M. Kimura, Y. Higuchi and H. Takagi, Appl.Phys. Express 2, 111401 (2009).19) L. Gao, H. Guo, S. Zhang and C. Randall, Actuators 5,8 (2016).20) T. B. Reed, “Free Energy of Formation of BinaryCompounds”, MIT Press, Cambridge, MA (1971).21) I. Karakaya and W. T. Thomspon, Bull. Alloy PhaseDiagr. 9, 237 (1988).22) V. Nguyen, H. Han, H. Lee, J. Yoon, K. Ahn and J. Lee,J. Ceram. Process. Res. 2, s282 (2012).23) N. Iwagami, H. Nagata, I. Sakaguchi and T. Takenaka,J. Ceram. Soc. Jpn. 124, 644 (2016).24) M. V. Slinkina, G. I. Dontsov and V. M. Zhukovsky,J. Mater. Sci. 28, 5189 (1993).25) H. Nagata, H. Haneda, I. Sakaguchi, T. Takenaka and J.Tanaka, J. Ceram. Soc. Jpn. 105, 805 (1997).26) C. Ahn, H. Song and S. Nahm, J. Am. Ceram. Soc. 89,921 (2006).27) S. Kakroo, A. Kumar, S. K. Mishra, V. Singh and P. K.Singh, “Phase Transit.”, Taylor & Francis (2016)p. 1063631.28) G. Yesner, J. Am. Ceram. Soc. 101, 5315 (2018).29) J. Lv, T. Karaki and M. Adachi, Jpn. J. Appl. Phys. 49,09MD06 (2010).30) G. Yesner, M. Kuciej and A. Safari (2015). doi:10.1109/ISAF.2015.717265831) M. Difeo, L. Ramajo and M. Castro, J. Adv. Dielectr. 11,2140004 (2021).32) H. Y. Tian, K. W. Kwok, H. L. W. Chan and C. E.Buckley, J. Mater. Sci. 42, 9750 (2007).33) F. Zhang, X. Qiao, Q. Shi, X. Chao, Z. Yang and D. Wu,J. Eur. Ceram. Soc. 41, 368 (2021).34) W. Jo, J. Ollagnier, J. Park, E. Anton, O. Kwon, C. Park,H. Seo, J. Lee, E. Erdem, R. Eichel and J. Rödel, J. Eur.Ceram. Soc. 31, 2107 (2011).35) C. Ahn, H. Kim, W. Woo, S. Won, H. Seog, S. Chae, B.Park, K. Jang, Y. Ok, H. Chong and I. Kim, J. Am.Ceram. Soc. 98, 1877 (2015).Fig. 10. Td as a function of the amount of Nb2O5 y.Journal of the Ceramic Society of Japan 133 [7] 321-328 2025 JCS-Japan32736) C. Chou, C. Liu, C. Hsiung and R. Yang, PowderTechnol. 210, 212 (2011).37) J. Kang, D. Heo, V. Nguyen, H. Han and J. Lee,J. Korean Phys. Soc. 61, 899 (2012).38) C. Lee, H. Han, S. Kim, T. Dinh, C. Ahn and J. Lee,J. Electroceram. 41, 4349 (2018).39) V. Schmitt and F. Raether, J. Eur. Ceram. Soc. 34, 1521(2014).40) T. Kujirai, Y. Takagi, H. Nagata and T. Takenaka(2019). doi:10.1109/ISAF43169.2019.903493641) H. Park, J. Choi, M. Choi, K. Cho and S. Nahm, J. Am.Ceram. Soc. 91, 2374 (2008).42) H. Han, J. Koruza, E. Patterson, J. Schulthei, E. Erdem,W. Joc, J. Leed and J. Rödel, J. Eur. Ceram. Soc. 37,2083 (2017).43) M. Matsubara, T. Yamaguchi, K. Kikuta and S.-I.Hirano, Jpn. J. Appl. Phys. 43, 7159 (2004).44) M. Matsubara, T. Yamaguchi, W. Sakamoto, K. Kikuta,T. Yogo and S.-I. Hirano, J. Am. Ceram. Soc. 88, 1190(2005).45) J. B. Lim, S. Zhang, J.-H. Jeon and T. R. Shrout, J. Am.Ceram. Soc. 93, 1218 (2010).46) D. Lin, K. W. Kwok and H. L. W. Chan, Appl. Phys.Lett. 90, 232903 (2007).47) S.-L. Yang, C.-C. Tsai, Y.-C. Liou, C.-S. Hong, B.-J. Liand S.-Y. Chu, J. Am. Ceram. Soc. 93, 1011 (2012).48) R. M. German, P. Suri and S. J. Park, J. Mater. Sci. 44,139 (2009).49) H. Wang, Y. Dai and X. Zhang, J. Am. Ceram. Soc. 95,1182 (2012).50) Q. Chen, L. Chen, Q. Li, X. Yue, D. Xiao and J. Zhu,J. Appl. Phys. 102, 104 (2007).51) C. Chou, C. Liu, C. Hsiung and R. Yang, PowderTechnol. 210, 212 (2011).52) M. Humenik and N. M. Parikh, J. Am. Ceram. Soc. 3,3960 (1956).53) Y. Imanaka, “Multilayered Low Temperature CofiredCeramics (LTCC) Technology”, Springer Science, Dor-drecht, The Netherlands (2005).54) K. Ojima, K. Iwasaki, Y. Takagi and H. Nagata,J. Ceram. Soc. Jpn. 131, 209 (2023).55) M. Li, M. J. Pietrowski, R. A. De Souza, H. Zhang,I. M. Reaney, S. N. Cook, J. Kilner and D. C. Sinclair,Nat. Mater. 13, 31 (2014).56) F. Lu, F. Fang and Y. Chen, J. Eur. Ceram. Soc. 21,1093 (2001).57) A. D. Pelton, Bull. Alloy Phase Diagr. 7, 2527 (1986).Iwasaki et al.: Electrical properties and 18O tracer diffusion in (Bi0.5Na0.5)TiO3 ceramics doped with CuO and Nb2O5JCS-Japan328